Why the Indian subcontinent holds the key to global tiger recovery QuickTime and a TIFF (Uncompressed) decompressor are needed to see this picture. Samrat Mondol K. Ullas Karanth Uma Ramakrishnan NCBS-TIFR CWS
Micro-evolutionary processes? Mutation migration drift
Current distribution Occupy ~ 7% of historical range ~ 3000-3500 tigers globally ~ 1500-2000 in India Largest diversity of habitats: Indian subcontinent What about genetic diversity?
Phylogeography Low diversity in India Sampling from one location Luo et al., 2004
Assessing tiger genetic variation in the Indian subcontinent QuickTime and a decompressor are needed to see this picture. Sampling: non-invasive scats from 73 tigers 28 protected areas including varied habitats Mondol et al., 2009
Assessing genetic variation: mitochondrial DNA H SNP =1 p i 2 n i=1 7 50 4 40 26 4 4 4 27 4 23 30 50 2 23 2 2 2 4 2 1 6 2 20 27 43 20 12 S ND1 ND2 CO1 ND5 ND6 Cyt B CR 0.22 0.33 0.15 0.16 0.15 0.11 0.25 0.27 9 20 8 2 44 20 20 26 4 10 10 43 13 2 20 4 2 30 26 15 26 44 44 20 50 20 131 bp 482 bp 450 bp 200 bp Ascertain for most variation Total sequence length: 1263 bp for 4 regions Mondol et al., PLoS gen 2009
Assessing genetic variation: nuclear DNA STRs: High mutation rate, very polymorphic, independently evolving, codominant loci Nuclear DNA from scat: degraded and low concentration 30 microsatellites from domestic cats, other tiger subspecies selected based on high heterozygosity and low allelic size range (<200 bp) Fecal DNA microsats: possible genotyping and amplification error All loci tested for amplification success with fecal DNA; 10 most consistent loci standardized; 5 of these loci used in Luo et al. for other subspecies Each locus genotyped 4 independent times for each sample. Final data includes samples with 75% or higher consistency Mondol et al., PLoS gen 2009
Genetic variation: mitochondrial DNA n= 68 32 haplotypes 76% genetic variation 10 haplotypes Resampling simulations reveal that Indian diversity is not higher due to sample s size Mondol et al., PLoS gen 2009
Genetic variation: nuclear microsatellites Subspecies Observed heterozygosity (S.D.) Number of alleles (S.D.) Allelic size range (S.D.) Bengal (P. tigris tigris) 0.70 (0.16) 12.4 (3.6) 32 (7.7) All other subspecies (Indo-Chinese, 0.53 (0.07) 7.2 (1.6) 16 (6.1) Malayan, Sumatran and Siberian) All South-East Asian subspecies 0.56 (0.14) 7.2 (1.6) 16 (6.1) (Indo-Chinese, Malayan and Sumatran) Indo-Chinese (P. tigris corbetti) 0.57 (0.27) 6.2 (1.5) 14.8 (4.8) Malayan (P. tigris jacksoni) and Sumatran (P. tigris sumatrae) 0.55 (0.05) 5.8 (1.5) 13.2 (6.1) India holds 63% of global genetic variation Mondol et al., PLoS gen 2009
Why are Indian tigers genetically more diverse? 1) Tigers originated in India 2) High Population differentiation 3) High Ancestral effective size
Indian origin for tigers? 1) Paleontological data suggest South China origin 2) Phylogenetic data Indian tigers Indo-chinese tigers Time 3) Population genetic models: LAMARC Indian tigers m12 m21 Other tigers Mt DNA: MLE (m21) = 185 (44, 486); MLE (m12) = 0.19 (0.01, 59) Nuclear DNA: MLE (m21) = 36 (31, 40); MLE (m12) = 13 (11, 15) Tigers expanded their range into India NO Mondol et al., PLoS gen 2009
High population differentiation? North (n=10) 2 Central (n=11) 2 South (n=18) 2 North (n=24) 1 0.027 (p=0.063) 0.041* (p=0.000) Central (n=18) 1 0.236* (p=0.000) 0.019 (p=0.054) South (n=26) 1 0.298* (p=0.000) 0.026 (p=0.279) High differentiation for mtdna South and central India not differentiated Structure contributes to high overall variation in Indian subcontinent Mondol et al., PLoS gen 2009
Fischer-Wright coalescent: Constant population size 1. Reconstruct genealogical relationship between the samples 2. Distribute mutations on the genealogy: number of mutations proportional to branch length Low S, π Quantifying genetic variation using statistics QuickTime and a Segregating sites: number of positions that differ TIFF (LZW) across decompressor samples Average pairwise difference: average number are needed of differences to see this between picture. pairs of sequences Constant size: seg sites = av. pairwise difference High S, π
Quantifying changes in population size with the coalescent QuickTime and a TIFF (LZW) decompressor are needed to see this picture. High S, low π Low S, high π
High Ancestral effective size? QuickTime and a TIFF (LZW) decompressor are needed to see this picture. QuickTime and a TIFF (LZW) decompressor are needed to see this picture. Peninsular Indian population reveals signal of decline Population decline quantified by other methods including LAMARC, m-ratio, BOTTLENECK Decline ~ 200 years old Mondol et al., PLoS gen 2009
How many tigers in Peninsular India? Effective population size: 23,280 (2,964, 151,008) Effective size / Census size = 0.4 ~ 58,202 adult tigers (7,412, 377,520) in peninsular India 200 years ago Given current estimates: decline of 98% Mondol et al., PLoS gen 2009
Sensitivity analyses Does magnitude of decline change when more genetic loci are used? NO Is our result valid only for peninsular Indian tigers? NO Mondol et al., PLoS gen 2009
Alternate explanations for high diversity in India Greater extent of population decline for other subspecies? QuickTime and a decompressor are needed to see this picture. QuickT TIFF (Uncompres are needed to NO Mondol et al., PLoS gen 2009
Conclusions Indian tigers have high genetic variation This high variation is due to population differentiation and high ancestral size However, we have already lost around 98% of these tigers
Implications for conservation The Indian subcontinent retains 50-60% of the global tiger population.. living in varied and fragmented habitats with 60-70% of species genetic variation.. Proportion of global tiger habitat in India: 8-25% Strong case for conservation of Indian tigers
Non-invasive genetic monitoring of tigers in Bandipur National Park Molecular methods to identify tigers (Mukherjee et al., 2007) Molecular sexing Genetic individual identification and population estimation Comparison to photographic mark-recapture estimates. QuickTime and a TIFF (Uncompressed) decompressor are needed to see this picture. Mondol et al., 2009 b 72 tiger scats 55 typed reliably at 5 loci 26 unique individuals (PID=0.005) Genetic population estimate= 66 (13) Photo-based estimate = 66 (13.8)
Future directions Sampling of historical skins to investigate lost variation, better quantify decline Quantifying phenotypic variation: striping pattern Landscape level studies in high tiger density areas to investigate connectivity
What drives patterns of genetic variation in the Indian subcontinent: Geography, Climate, Ecology or Humans? Uma Ramakrishnan
Global drivers of patterns of genetic diversity? Biogeographic divides Recent climatic fluctuations Ecology: dispersal ability and population size
Why the Indian subcontinent? Three major biogeographic realms (Palearctic, Africotropical, Indomalayan) intersect here Geologically interesting history Ecologically encompasses a diversity of habitat types Hominins have been present since the last million years Data poor
Where is the diversity in the Indian subcontinent from? Palearctic tigers African Leopards Jungle cat? India part of seconday range expansion. Oriental Sinica macaques Small-clawed and smooth-coated otters Leopard cat
In the Indian subcontinent No major biogeographic divides Large differences in elevation across the subcontinent Impacts of climate not very clearly understood: Posters: Robin, Priya Ecology: different patterns for very large and very small species Significant anthropogenic impacts
How do we test these predictions? Comparative framework Do differences in elevation matter? Contrast species living in high elevations with closely related species in plains, Arunachal vs bonnet macaques Do differences in body size matter? Contrast species across a range of body sizes, tigers, leopards, jungle cats Do differences in climatic regime matter? Contrast ecologically similar species with small differences in habitat preference, leopard cat vs jungle cat Do anthropogenic impacts matter? Contrast species that have been impacted negatively by humans (tigers) with unknown impacts (leopards)
Genetic patterns in two macaques Largest primate genus Most widespread primate (after humans) Most diverse distribution (after humans) Discovered in 2005 Identified as a genetically distinct species by us in 2007 Very high morphological and behavioural variation species to individual Highly adaptable to diverse environments
QuickTime and a 3000m Munzala phylogenetic tree and network 2000m West Siang Upper Subansiri 325 km Mitochondrial DNA Tawang Arunachal Pradesh 80 or more BS 50 or more BS 0.77* decompressor are needed to see this picture. 0.76* 0.76* Strong signal of population differentiation in Munzala
Biological Sampling 1540km n= 38
>70% Bootstrap network
Bonnet macaque phylogenetic tree and network Much less genetic differentiation than Munzala A hint that the Palghat gap might be a biogeographic divide
Ecological parameters scale with body size Carbone & Gittleman, 2002 Sutherland et al., 2002
Differences in body size relative density relative dispersal 35 30 25 20 15 High density Low movement intermediate density, movement Low density high movement 10 5 0 Jungle cat Leopard Tiger 3 kg 40 kg 150 kg
How do differences in density, movement impact genetic structure? N e m N e ~ Density/local population size m ~ dispersal m/dispersal Low N e High m Low variation low differentiation Tiger? Low N e Low m Low variation high differentiation Leopard? High N e High m high variation low differentiation High N e high variation moderate differentiation Jungle Low cat? m N e /Density
Differences in body size: isolation by distance across peninsular India 3,000 km Low intercept: low variation Low slope: low differentiation v v 3,000 km n = 73 n = 88 n = 55 Sampling
Differences in body size: isolation by distance across peninsular India 0.008000 0.007000 y = 1E-06x + 0.0042 R 2 = 0.7201 0.006000 High intercept: high variation Intermediate slope 0.005000 y = 1E-06x + 0.0025 R 2 = 0.882 0.004000 0.003000 Int intercept: moderate variation Slope higher than tiger y = 5E-07x + 0.0019 R 2 = 0.492 0.002000 0.001000 Low intercept: low variation Low slope: low differentiation 0.000000 0 500 1000 1500 2000 2500 3000 geographic distance Continuously distributed species reveal genetic patterns driven by body siz
Same body size, but different origins Mukherjee et al., PLoS One, in revision
Length of tail as a percentage of head + body length in some cats 110 Length of tail % of head + body length 100 90 80 70 60 50 40 30 20 Marbled Cat n=3 Clouded Leopard n= 6 Asian origin Asiatic Golden Cat n=3 Means with ranges Leopard Cat n=10 Rusty spotted Cat n=8 Black Footed Cat African origin Jungle Cat n=44 Serval Caracal n = 5 Pocock (1939), http://www.abf90.dial.pipex.com/bco/ver4.htm Mukherjee et al., PLoS One, in revision
QuickTime and a decompressor are needed to see this picture. Phylogenetic trees and haplotype networks 543 scats from various biogeographic zones Used PCR-RFLP to ascertain jungle cats and leopard cats Jungle cat: 55; Leopard cat: 40 Mukherjee et al., PLoS One, in revision NADH5: 460bp jungle cat, 362bp leopard cat Cytochrome b: 141bp jungle cat, 202 bp leopard cat
F. c. valbalala F. c. kutas F. c. affinis Population subdivision Taxonomy (n) F. c. valbalala (13) F. c. kutas (23) 0.04 F. c. affinis (10) 0.20 0.12 P. b. horsfieldi P. b. bengalensis 0.32 F. c. prateri (9) 0.14 0.07 0.16 Latitudinal range (n) 10-19.9 20-28.9 10-19.9 (15) 20-28.9 (30) 0.05 29-35 (10) 0.19 0.12 Biogeographic zones Himalaya N. East N. East (9) 0.30 W Ghats (12) 0.90 0.91 Jungle cat patterns are as expected Continuous distribution throughout India Weak isolation by distance Leopard cat patterns in contrast to expectation A clear break in geographical continuity indicated by genetic differentiation Are we missing populations in-between in Central India? If there is a gap in distribution, what is causing it? Mukherjee et al., PLoS One, in revision
Are we missing populations in-between in Central India? If there is a gap in distribution, what is causing it? Niche model analysis for leopard cat using climatic variables (temperature and rainfall) Mukherjee et al., PLoS One, in revision
Leopard cat locations superimposed over the maximum temperatures in the warmest month. Mukherjee et al., PLoS One, in revision Mean = 29.27 o C (95% CI: 28.59 o C - 29.93 o C; n = 217)